The University of New South Wales School of Computer Science & Engineering
Road to RoboCup 2006 rUNSWift 2006 Behaviors & Vision Optimizations
by Enyang HUANG
Submitted as a requirement for the degree Bachelor of Software Engineering with Honors Submitted: 1 September 2006 Supervisor: Dr William Uther Assessor: Professor Claude Sammut
1
Abstract This thesis describes work done in visions and high-level behaviors in 2006 at rUNSWift1 . For both parts, we have introduced new approaches and re-implemented bunch of the modules. The rst part of this thesis, Vision, covers the 2006 implementation of chromatical correction of raw images with machine learning approach. Moreover, as a result of the new rule on the eld, we also introduced the concept of eld edge, which is used by higher level agent vision processes. The second part of this thesis is a review of the 2006 rUNSWift behavior system. Lots of the modules and concepts here are, re-dened carefully and re-implemented, with concentration from simple low-level skills such as ball tracking, ball nding, ball grabbing, ball dribbling, ball kicking, to high-level complex behaviors such as role assignments, positioning and team cooperation. In parallel, a strategy simulator was implemented which is capable of performing plug-in-and-play simulation of the behavior modules. The team comes 1st in Australian Open 20062 , and 2nd in Robocup 2006 Bremen3 , using 2006 rUNSWift code base.
rUNSWift is a combination entry of the University of New South Wales and Nation ICT Australia in Robocup 4 legged-league. National ICT Australia is funded through the Australian Government's Backing Australia's Ability initiative, in part through the Australian Research Council. 2 See http://www.cse.unsw.edu.au/robocup 3 See http://www.robocup2006.org
1
Contents
1 Introduction 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 rUNSWift 2006 Architecture . . . . . . . . . . . . . . . . . . 1.3 Report Overview . . . . . . . . . . . . . . . . . . . . . . . . . 2 Ring Correction 2.1 Overview . . . . . . . . . . . 2.2 Past Work . . . . . . . . . . . 2.3 Presentation of the work . . . 2.3.1 Theory . . . . . . . . 2.3.2 Function Minimization 2.4 Implmentation . . . . . . . . 2.5 Summary . . . . . . . . . . . 8 8 8 11 12 12 12 13 14 15 16 18 20 20 22 22 23 24 25 28 30 30 30 32 34 34 38 40 45 45

下一页